We use experimental and survey data from two natural-hazard prone countries in Latin America to test the hypothesis that natural hazards affect risk aversion. We use two methods to measure risk aversion: simple questions on the willingness to pay for a hypothetical lottery and more complicated experiments involving real pay-offs. We find that whereas the experiments provide reasonable estimates of risk aversion, the hypothetical questions result in unrealistic distributions of preferences. The experimental results strongly support the hypothesis that experiencing natural shocks makes people more risk averse, not only in the short run but also in the medium and long run.